| 0850 | Turning on FET heaters in DIB1/DCRC31 causes temperature to fall to 36 mK (still high-ish), charge channels to return but just choppy noise. Efforts to un-rail a phonon channel fail. |
| 0855 | Fill LN cold trap Dewar |
| 0905 | MC/Tower = 32.1/37 mK, Cold plate = 188 mK, Still = 2 mW / 1.53 K / 0.140 mbar, G1/G2 = 225/210 mbar, 1K Pot = 1.70 K / 18 mbar, Main bath = 34.8%. I'm a bit concerned about the underlined items… |
| 0910 | Power down all three DCRCs. Still P drops, MC temperature falling slowly to previous base. |
| 0949 | MC/Tower = 24.1/30.7 mK, Cold plate = 134 mK, Still = 2 mW / 1.05 K / 0.140 mbar, G1/G2 = 200/185 mbar, 1K Pot = 1.69 K / 17 mbar. This is much better, back to normal. Were the DCRCs heating up the still through the rod? If something was seriously wrong at the SQUID card this may explain both the loss of phonon channels and the elevated Still and CP temperatures. |
| 1005 | Find that snowdon is full and we cannot copy new data to it. For time being, copy all Run 53 data to /data/chocula/k100/raw/byseries instead. |
| 1030 | Data taken last night performing a sweep of SQUID Lockpoints was found to be bad, with several railed phonon channels on all dibs, and charge channels on dib 1 showing noise but no pulses (perhaps due to not typing wr 33 2000 before data taking started). |
| 1110 | Conduct study to determine whether flashing the detectors was responsible for altering the proper driver offset value and thus railing out phonon channels during data taking. Set appropriate values on all channels such that pulses look acceptable, then flash for roughly 6 seconds (while setting Qinner and Qouter to 0V). We found that our first flashing attempt set the tower temperature to ~45 mK, not sufficient for a flashing. Proceeding attempts at flashing did not increase the fridge temperature at all from the base of 31 mK, though turning on the LEDs did show oscillatory patterns in pulses. |
| 1145 | For no known reason, flashing eventually functioned as expected after several attempts. Reset boards such that pulse look good, and flash again for several seconds (reaching tower temperature of 160 mK). After 15 minutes of cooldown time (tower temperature of 32 mK), pulses on all three dibs look good, with no railed channels. It appears flashing is not responsible for railed channels during data taking. |
| 1550 | Commence LHe transfer (21.8%) |
| 1624 | Finish LHe transfer (94.0%) |
| 1742 | Struggling to get DAQ in good state to take simple background data. Modified vi's so that SQUIDS could be relocked before run. DIB1 went into bad state after flashing. Couldn't get charge pulses or unrail phonons. Powercycled DCRC31. Temp was high but finally cooled off. Got phonons back but had to set QET biases to only about 2/3 their previous good values. Meanwhile DIB2 has two railed channels. These were not difficult to reset. DIB3 has no problems. |
| 1802 | Giving up - pretty sure we need to power cycle DCRC31 again but we don't have time. Will flash detector overnight. |